1. Field of the Invention
Embodiments relate to a digital image signal processing apparatus and method capable of generating a data patch for noise reduction filtering, performing a high-speed operation, and thus generating an output image having the same size as an input image while maintaining the same quality as the input image.
2. Description of the Related Art
With respect to digital image signal processing apparatuses, megapixel-related competition is in saturation and thus competition between manufacturers has expanded to high sensitivity photographing. In this regard, there is a problem since noise due to heat generated by a charge coupled device (CCD) or a complementary metal-oxide semiconductor (CMOS), or low frequency noise due to interference in an electric circuit is intensified due to signal amplification when the high sensitivity photographing is performed. Thus, efficient reduction of noise generated in an output image becomes an important issue.
According to conventional technology, a low pass filter (LPF) is often used to simply perform noise reduction (NR). Via an LPF, since a noise component is generated as a high frequency component in a plane portion of an image, noise is reduced by cutting off the high frequency component.
However, by using an LPF, an edge component of an image may also be reduced such that sharpness of the image deteriorates, and in this regard, there is a demand for technology capable of only reducing noise while maintaining edge components.
In this regard, technology for performing NR by sampling an image and then processing the image via software is introduced.
However, such a NR technology based on software processing has a disadvantage in that high-speed processing is impossible. That is, all calculations are performed in a central processing unit (CPU) and thus, processing speed is dependent upon the CPU. However, the CPU has to perform not only image processing but also performs other operations simultaneously, and thus, the processing speed is limited. In addition, image data before being processed via software is stored in a storage device including a memory, secure digital (SD)/compact flash (CF)/smart media (SM) cards or the like, and speed for retrieving the image data stored in the storage device for a process via the software is considerably limited.